Nearest Neighbor Speculative Decoding for LLM Generation and Attribution

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Li, Minghan, Chen, Xilun, Holtzman, Ari, Chen, Beidi, Lin, Jimmy, Yih, Wen-tau, Lin, Xi Victoria
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916706886090752
author Li, Minghan
Chen, Xilun
Holtzman, Ari
Chen, Beidi
Lin, Jimmy
Yih, Wen-tau
Lin, Xi Victoria
author_facet Li, Minghan
Chen, Xilun
Holtzman, Ari
Chen, Beidi
Lin, Jimmy
Yih, Wen-tau
Lin, Xi Victoria
contents Large language models (LLMs) often hallucinate and lack the ability to provide attribution for their generations. Semi-parametric LMs, such as kNN-LM, approach these limitations by refining the output of an LM for a given prompt using its nearest neighbor matches in a non-parametric data store. However, these models often exhibit slow inference speeds and produce non-fluent texts. In this paper, we introduce Nearest Neighbor Speculative Decoding (NEST), a novel semi-parametric language modeling approach that is capable of incorporating real-world text spans of arbitrary length into the LM generations and providing attribution to their sources. NEST performs token-level retrieval at each inference step to compute a semi-parametric mixture distribution and identify promising span continuations in a corpus. It then uses an approximate speculative decoding procedure that accepts a prefix of the retrieved span or generates a new token. NEST significantly enhances the generation quality and attribution rate of the base LM across a variety of knowledge-intensive tasks, surpassing the conventional kNN-LM method and performing competitively with in-context retrieval augmentation. In addition, NEST substantially improves the generation speed, achieving a 1.8x speedup in inference time when applied to Llama-2-Chat 70B. Code will be released at https://github.com/facebookresearch/NEST/tree/main.
format Preprint
id arxiv_https___arxiv_org_abs_2405_19325
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Nearest Neighbor Speculative Decoding for LLM Generation and Attribution
Li, Minghan
Chen, Xilun
Holtzman, Ari
Chen, Beidi
Lin, Jimmy
Yih, Wen-tau
Lin, Xi Victoria
Computation and Language
Large language models (LLMs) often hallucinate and lack the ability to provide attribution for their generations. Semi-parametric LMs, such as kNN-LM, approach these limitations by refining the output of an LM for a given prompt using its nearest neighbor matches in a non-parametric data store. However, these models often exhibit slow inference speeds and produce non-fluent texts. In this paper, we introduce Nearest Neighbor Speculative Decoding (NEST), a novel semi-parametric language modeling approach that is capable of incorporating real-world text spans of arbitrary length into the LM generations and providing attribution to their sources. NEST performs token-level retrieval at each inference step to compute a semi-parametric mixture distribution and identify promising span continuations in a corpus. It then uses an approximate speculative decoding procedure that accepts a prefix of the retrieved span or generates a new token. NEST significantly enhances the generation quality and attribution rate of the base LM across a variety of knowledge-intensive tasks, surpassing the conventional kNN-LM method and performing competitively with in-context retrieval augmentation. In addition, NEST substantially improves the generation speed, achieving a 1.8x speedup in inference time when applied to Llama-2-Chat 70B. Code will be released at https://github.com/facebookresearch/NEST/tree/main.
title Nearest Neighbor Speculative Decoding for LLM Generation and Attribution
topic Computation and Language
url https://arxiv.org/abs/2405.19325